IBS-D symptom dynamic monitoring and doctor-patient interaction terminal

Abstract

The present application relates to the technical field of medical information analysis, in particular to a IBS-D symptom dynamic monitoring and doctor-patient interaction terminal, the present application extracts first characteristic medical information and second characteristic medical information through an irritable bowel feedback analysis model, and the second characteristic medical information is calibrated and quantified to obtain third characteristic medical information under the condition of containing the first characteristic medical information; the first characteristic medical information and the third characteristic medical information are combined as key quantitative information to generate an IBS-D medical monitoring feature report; through the method, the physiological information and the communication information of the monitored IBS-D patient can be quantified, and a medical monitoring feature report is generated based on the information, so as to expand the application range of the monitoring data to the treatment data, reduce the communication information error caused by the disease of the patient, and provide accurate data for the doctor to make data reference.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of medical information analysis, in particular to an IBS-D symptom dynamic monitoring and doctor-patient interaction terminal. BACKGROUND

[0002] Diarrhea-predominant Irritable Bowel Syndrome (IBS-D) is a subtype of Irritable Bowel Syndrome (IBS) and a gastrointestinal disease. In addition to common gastrointestinal disease symptoms such as abdominal pain, abdominal distension, and diarrhea, IBS-D patients usually also exhibit anxiety and depression. Anxiety and depression of patients can also aggravate the condition of IBS-D patients (gut-brain interaction). Therefore, in addition to normal drug treatment, emotional comfort for IBS-D patients is also one of the current treatment measures.

[0003] In a conventional treatment scheme for IBS-D patients, doctors usually detect treatment effect and perform emotional comfort through active outpatient follow-up of patients. However, there are problems such as inability to provide timely feedback, high loss to follow-up, and serious consumption of medical resources. In addition, there are the following problems in relying on internet communication: there is a large subjectivity in describing symptoms and the like, and objective physiological indicators of patients cannot be reflected. Therefore, there is a lack of a more efficient treatment feedback mode between doctors and patients. With the development of technology, it is possible to remotely monitor various physiological information of IBS-D through an intestinal sound monitoring device and the like, to remotely communicate with IBS-D patients, to reasonably adjust medication, and to comfort patients. Remote treatment of IBS-D can be achieved, and medical resources can be saved to the greatest extent.

[0004] However, there are still some problems in the treatment process. On the one hand, the main current intestinal sound monitoring device is disposable and has poor endurance, so the feedback monitoring data is limited, and the feedback of treatment data has limitations. On the other hand, due to the anxiety and depression of IBS-D patients, the information conveyed by IBS-D patients in the communication process is also affected by emotions and is inaccurate. In order to prevent IBS-D patients from being affected by emotions and causing the condition to worsen, doctors cannot ask questions as they would for other diseases, resulting in a large error in the feedback communication information data, and the accuracy of the feedback of treatment data is also affected to a certain extent.

[0005] Therefore, an IBS-D symptom dynamic monitoring and doctor-patient interaction terminal is provided. SUMMARY

[0006] The present invention aims to provide an IBS-D symptom dynamic monitoring and doctor-patient interaction terminal. The terminal extracts first characteristic medical information and second characteristic medical information through an irritable bowel feedback analysis model, performs data quantification and calibration on the second characteristic medical information when the first characteristic medical information is included, and obtains third characteristic medical information. The first characteristic medical information and the third characteristic medical information are combined as key quantitative information to generate an IBS-D medical monitoring characteristic report. This method can quantify the physiological information and communication information of monitored IBS-D patients, and generate a medical monitoring characteristic report based on this information, thereby expanding the applicability of monitoring data to treatment data, reducing communication information errors caused by symptoms, and providing doctors with accurate treatment feedback data for data reference.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] An IBS-D symptom dynamic monitoring and doctor-patient interaction terminal, comprising:

[0009] A medical information collection module is configured to obtain first medical information and second medical information of an IBS-D patient; the first medical information includes vital sign information and physiological activity information acquired by a medical device; the second medical information includes voice communication information and text communication information of the IBS-D patient;

[0010] A medical information analysis module transmits the first medical information to an irritable bowel feedback analysis model, determines whether there is first characteristic medical information of the first medical information, and if so, extracts the first characteristic medical information and stores the data;

[0011] Furthermore, the first characteristic medical information is obtained through the abnormal irritable bowel physiological characteristic model of the irritable bowel feedback analysis model;

[0012] The abnormal irritable bowel physiological characteristic model includes a physiological characteristic data processing unit, a physiological characteristic data extraction unit, a physiological characteristic abnormality calculation unit and a physiological abnormality characteristic output unit;

[0013] The physiological characteristic data processing unit performs data preprocessing on the first medical information;

[0014] The physiological characteristic data extraction unit extracts IBS-D medical information features from the preprocessed first medical information; the physiological characteristic data extraction unit includes 5 LSTM layers, 2 spatiotemporal convolution layers, 1 attention mechanism layer, and 1 feature activation layer; the physiological characteristic abnormality calculation unit performs feature similarity based on the IBS-D medical information features to determine whether the first feature medical information exists; specifically:

[0015] σ IBS-D =fIBS-D (d F )·ω1·Cos(d F ,d U )+f IBS-D (d F )·ω2·Dtw(d F ,d U );

[0016] Among them, σ IBS-D is the similarity of the first feature medical information, f IBS-D (d F ) is the IBS-D medical information feature extracted from the first medical information; Cos is the cosine similarity algorithm, d F For the first medical information, d U is the abnormal medical information of IBS-D, Dtw is the dynamic time planning algorithm, ω1 and ω2 are the feature similarity weights;

[0017] The physiological abnormality feature output unit outputs the first feature medical information and stores the data;

[0018] transmitting the second medical information to the irritable bowel feedback analysis model, performing data comparison between the second medical information and stored historical second medical information, generating second characteristic medical information and storing the data;

[0019] Furthermore, the second characteristic medical information is obtained based on the medical feedback information analysis model of the irritable bowel feedback analysis model;

[0020] The medical feedback information analysis model includes a medical feedback information processing unit, a medical feedback key information extraction unit and a medical feedback weight calculation and output unit;

[0021] The medical feedback information processing unit performs data feature processing on the second medical information;

[0022] The medical feedback key information extraction unit compares the second medical information after data feature processing with the historical second medical information to extract key semantic features, and the medical feedback key information extraction unit adopts a Transformer model;

[0023] The medical feedback weight calculation and output unit performs feedback key weight calculation based on the key semantic features to generate second feature medical information; specifically:

[0024]

[0025] in, is the i-th feedback key weight, i∈(1,N j ), Nj the second medical information total number of the jth cluster category, j e (1, n), n is the total number of cluster categories; ln is the logarithmic function with e as the base, Num is the number of key semantic feature statistics; Tran is the Transformer model, the lth second medical information, l e (1, M), M is the total number of second medical information, Km is the semantic clustering feature label data, β1 and β2 are feedback key weight parameters;

[0026] The medical information association module associates the second characteristic medical information with the first characteristic medical information to obtain third characteristic medical information if the first characteristic medical information exists, and stores the third characteristic medical information; specifically: Wherein, the jth third characteristic medical information, i e (1, N j ), j e (1, n);

[0027] If the first characteristic medical information does not exist, the second characteristic medical information is stored as the third characteristic medical information;

[0028] The medical monitoring feature report output module outputs the first characteristic medical information and the third characteristic medical information to generate an IBS-D medical monitoring feature report if the first characteristic medical information exists, and outputs the third characteristic medical information to generate the IBS-D medical monitoring feature report if the first characteristic medical information does not exist.

[0029] Further, the initial report of the IBS-D medical monitoring feature report includes one IBS-D medical abnormal information label and K IBS-D medical conversation information labels, K>1; each IBS-D medical conversation information label corresponds to k third characteristic medical information, k>1; and the IBS-D medical abnormal information label corresponds to the first characteristic medical information.

[0030] If the first characteristic medical information is not less than the abnormal medical information threshold, the initial report outputs the IBS-D medical abnormal information label; and if the sum of the k third characteristic medical information is not less than the medical conversation information threshold, the initial report outputs the corresponding IBS-D medical conversation information label.

[0031] The IBS-D medical monitoring feature report is generated according to the output IBS-D medical abnormal information label and the output IBS-D medical conversation information label.

[0032] Compared with the prior art, the present application has the following advantages:

[0033] 1. Based on the first and second medical information of IBS-D patients, the irritable bowel feedback analysis model is used to analyze whether the first characteristic medical information exists, and the second characteristic medical information is extracted based on the irritable bowel feedback analysis model and the first characteristic medical information. This method can mine the deep features of the physiological information monitored by IBS-D patients based on existing data, expand the scope of use of the monitored physiological information, and also extract key semantic information of IBS-D patients based on the physiological information, reduce the error of communication monitoring data, provide an accurate data basis for subsequent treatment data feedback, and facilitate data reference for doctors.

[0034] 2. Based on the deep features of IBS-D patients' monitored physiological information and their key semantic information, the feedback key weights are calculated through a clustering method. This method can further expand the data usage scope of the monitored physiological information and also screen the semantic features of patients based on the monitored physiological information. This method can quantify various data, facilitate the calculation and feedback of subsequent treatment data, provide a quantitative data foundation for doctors' subsequent work, and facilitate data reference for doctors.

[0035] 3. The present invention uses the deep monitoring characteristics of the first characteristic medical information after data quantification and the third characteristic medical information after data quantification after screening, and converts the quantified data into an IBS-D medical monitoring characteristic report based on the initial report. This method can minimize the errors in communication information data caused by emotional issues of IBS-D patients, provide an accurate data basis for subsequent treatment data feedback, and facilitate data reference for doctors. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 This is a flow chart of the terminal system of the present invention;

[0037] Figure 2 This is a terminal Internet of Things diagram of the present invention;

[0038] Figure 3 A virtual device diagram of the abnormal irritable bowel physiological characteristic model and the medical feedback information analysis model of the irritable bowel feedback analysis model of the present invention;

[0039] Figure 4 This is a network structure diagram of the physiological characteristic data extraction unit of the present invention.

[0040] In the figure: 1. IBS-D patient; 2. Sensor; 3. Server; 4. Mobile terminal; 5. Network; 6. Doctor. DETAILED DESCRIPTION

[0041] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0042] In addition to gastrointestinal symptoms, IBS-D patients are also prone to anxiety and depression, which can also affect the symptoms of IBS-D. Therefore, in addition to drug treatment, doctors also need to comfort patients emotionally. However, there are still some problems in the treatment process. First, the monitoring instrument cannot monitor the patient in real time for a long time, resulting in certain limitations in the feedback of the monitoring data on the treatment. At the same time, the patient's emotions will also affect the information communicated between the doctor and the patient, resulting in errors in the communication information and affecting the feedback of the treatment data. For this reason, the present invention provides an IBS-D symptom dynamic monitoring and doctor-patient interaction terminal, reference Figure 1 and Figure 2 The technical solution shown is as follows:

[0043] Medical information collection module, obtains the first medical information and the second medical information of IBS-D patients; corresponds to Figure 2 In the process, the server 3 obtains the first medical information of the IBS-D patient 1 collected by the sensor 2 through the network 5; the server 3 obtains the second medical information of the IBS-D patient 1 collected by the mobile terminal 4 through the network 5;

[0044] The medical information analysis module transmits the first medical information to the irritable bowel feedback analysis model, determines whether there is the first characteristic medical information of the first medical information, and if so, extracts the first characteristic medical information and stores the data; transmits the second medical information to the irritable bowel feedback analysis model, compares the second medical information with the stored historical second medical information, generates the second characteristic medical information and stores the data; the medical information analysis module Figure 2 Completed in server 3;

[0045] The medical information association module, if the first characteristic medical information exists, obtains the third characteristic medical information based on the association of the second characteristic medical information with the first characteristic medical information, and stores the third characteristic medical information; if the first characteristic medical information does not exist, stores the second characteristic medical information as the third characteristic medical information; the medical information association module Figure 2 Completed in server 3;

[0046] a medical monitoring feature report output module, if the first feature medical information exists, outputting the first feature medical information and the third feature medical information to generate an IBS-D medical monitoring feature report; if the first feature medical information does not exist, outputting the third feature medical information to generate the IBS-D medical monitoring feature report; corresponding Figure 2 Specifically, the server 3 analyzes and transmits the IBS-D medical monitoring feature report to the doctor 6 through the network 5.

[0047] The first feature medical information of the first medical information of the IBS-D patient is analyzed and judged by establishing an intestinal irritability feedback analysis model, and the existing first feature medical information is extracted. The second medical information of the IBS-D patient is screened in combination with the historical second feature medical information. Based on the above process, the deep data basis of the existing monitoring data of the IBS-D patient can be mined, the monitoring range of the data depth is improved, and more rich monitoring data basis is provided for subsequent data feedback. At the same time, by comparing with the historical second feature medical information, the semantic information of the patient affected by the emotion can be preliminarily screened, the semantic information error caused by the emotion is reduced, the accuracy of the communication information data is improved, and an accurate data basis is provided for subsequent data feedback, which is convenient for doctors to refer to data.

[0048] On this basis, the second feature medical information is further screened based on the first feature medical information to generate the third feature medical information, and the IBS-D medical monitoring feature report is generated in combination with the first feature medical information and the third feature medical information. Through this step, the physiological feature information and the semantic feature information can be quantified, and the corresponding weight can be generated. Through the weight, the initial report content in the IBS-D medical monitoring feature report can be calculated, and the corresponding medical monitoring feature report data can be obtained according to the weight calculation result. Based on this method, the actual data can be uniformly quantized, and based on the quantized data, the feedback work related data calculation of the subsequent IBS-D medical monitoring feature report can be facilitated. Through these data, the communication information error caused by the emotional problem of the IBS-D patient can be reduced, and the data use range of the physiological monitoring data can be expanded, which provides an accurate data basis for subsequent data feedback and is convenient for doctors to refer to data.

[0049] Embodiment one

[0050] In order to illustrate specifically, the following content is combined for explanation:

[0051] A medical information collection module acquires first and second medical information of the IBS-D patient; the first medical information includes vital sign information (such as, but not limited to, heart rate, blood pressure, etc.) and physiological activity information (such as, but not limited to, bowel sounds, etc.) acquired by medical equipment, and the first medical information is acquired through relevant medical monitoring equipment; the second medical information includes voice communication information and text communication information (such as, but not limited to, WeChat, etc.) of the IBS-D patient;

[0052] During the actual treatment process of IBS-D patients, in addition to monitoring the patient's condition through various physiological activities and vital signs to provide reference data for drug treatment, chat data is also a reference for IBS-D patient treatment, as IBS-D patients are prone to emotional problems and their symptoms may be affected. Therefore, obtaining these two types of comprehensive data will help to comprehensively analyze the subsequent data, provide an accurate data foundation for subsequent data feedback, and facilitate data reference for doctors.

[0053] A medical information analysis module transmits the first medical information to an irritable bowel feedback analysis model, determines whether there is first characteristic medical information of the first medical information, and if so, extracts the first characteristic medical information and stores the data;

[0054] Furthermore, the first characteristic medical information is obtained through the abnormal irritable bowel physiological characteristic model of the irritable bowel feedback analysis model;

[0055] The abnormal irritable bowel physiological characteristic model includes a physiological characteristic data processing unit, a physiological characteristic data extraction unit, a physiological characteristic abnormality calculation unit and a physiological abnormality characteristic output unit. Figure 3 The abnormal physiological characteristics of irritable bowel are shown in the model;

[0056] The physiological characteristic data processing unit performs data preprocessing on the first medical information; the physiological characteristic data extraction unit extracts IBS-D medical information features from the preprocessed first medical information; the physiological characteristic data extraction unit includes 5 LSTM layers, 2 spatiotemporal convolution layers, 1 attention mechanism layer, and 1 feature activation layer; the physiological characteristic abnormality calculation unit performs feature similarity based on the IBS-D medical information features to determine whether the first feature medical information exists; specifically:

[0057] σ IBS-D =f IBS-D (d F )·ω1·Cos(d F ,d U )+f IBS-D (d F )·ω2·Dtw(dF ,d U );

[0058] Among them, σ IBS-D is the similarity of the first feature medical information, f IBS-D (d F ) is the IBS-D medical information feature extracted from the first medical information; Cos is the cosine similarity algorithm, d F For the first medical information, d U is the abnormal medical information of IBS-D, Dtw is the dynamic time planning algorithm, ω1 and ω2 are the feature similarity weights; d U In order to combine the abnormal historical physiological information and abnormal historical vital sign information in the historical IBS-D patient monitoring data extracted and stored by experts; ω1 and ω2 are set to 0.5 and 0.5 by default, and can be changed according to actual conditions; f IBS-D (d F ) takes a value of 0 or 1, 0 represents no abnormality, and 1 represents an abnormality;

[0059] By mining the abnormal features in the deep-level data characteristics of various physiological data and vital signs of IBS-D patients as data weights for data quantification, and using the abnormal physiological data and abnormal vital signs data related to IBS-D as the data foundation and performing similarity comparisons in data morphology, the scope of use of monitoring data can be expanded, the limitations of current monitoring data usage can be alleviated, and a data foundation can be provided for subsequent IBS-D medical monitoring feature reporting and information communication analysis, making it easier for doctors to use data for reference.

[0060] The physiological abnormality feature output unit outputs the first feature medical information and stores the data;

[0061] transmitting the second medical information to the irritable bowel feedback analysis model, performing data comparison between the second medical information and stored historical second medical information, generating second characteristic medical information and storing the data;

[0062] Furthermore, the second characteristic medical information is obtained according to the medical feedback information analysis model of the irritable bowel feedback analysis model; the medical feedback information analysis model includes a medical feedback information processing unit, a medical feedback key information extraction unit and a medical feedback weight calculation output unit, referring to Figure 3 If the information is text communication, it is directly processed; if the information is voice communication, the voice is converted into text, including but not limited to the voice-to-text function of the chat communication platform, or open source speech recognition models such as Whisper and Qwen2;

[0063] By extracting semantic features from the communication data of IBS-D patients, we can identify key semantic features in the data. This method can significantly reduce the error in communication data caused by emotional issues in IBS-D patients. It can also provide data reference for subsequent feedback through key semantic features.

[0064] The medical feedback information processing unit performs data feature processing on the second medical information; the data feature processing includes data format unification, punctuation data cleaning and dictionary-based maximum forward matching algorithm;

[0065] The medical feedback key information extraction unit compares the second medical information after data feature processing with the historical second medical information to extract key semantic features, and the medical feedback key information extraction unit adopts a Transformer model; the historical second medical information is semantic feature data of the current patient and other patients stored in combination with expert experience when they are not affected by the emotions of IBS-D symptoms;

[0066] The medical feedback weight calculation and output unit performs feedback key weight calculation based on the key semantic features to generate second feature medical information; specifically:

[0067]

[0068] in, is the i-th feedback key weight, i∈(1,N j ), N j is the total number of the second medical information of the jth cluster category, j∈(1,n), n is the total number of cluster categories; ln is the logarithmic function with e as the base, Num is the number of key semantic features; Tran is the Transformer model, is the lth second medical information, l∈(1,M), M is the total number of second medical information, Km is the semantic clustering feature label data, β1 and β2 are the feedback key weight parameters; β1+β2=1;

[0069] The clustering method specifically involves obtaining key semantics (such as, but not limited to, good, good, and bad) from historical communication data of IBS-D patients. Based on expert experience, the semantic representations are assigned a default range of 0-5, with 0 representing good and 5 representing bad. This range can be modified based on actual conditions. A corresponding weight is set for each key semantic. The weight represents the default relevance to IBS-D symptoms, ranging from 0-10, with 0 representing no relevance and 10 representing complete relevance. This range is not unique and can be modified based on actual conditions. The data weight of each semantic is specifically set based on expert experience.

[0070] A two-dimensional coordinate system is established with semantic weight as the horizontal axis and semantic quality as the vertical axis. All key semantics are generated based on the data corresponding to the weight and quality and filled into the two-dimensional coordinate system. Clustering is performed according to the k-means clustering algorithm. The value of k is set based on expert experience. The center coordinates of each cluster range after clustering are used as the semantic weight and semantic quality of the cluster. The semantic cluster feature label data of each cluster range is calculated as follows:

[0071]

[0072] Among them, δ k-means is the semantic clustering feature label data of each cluster range, δ x is the abscissa of the cluster center, δ y is the vertical coordinate of the cluster center, is the maximum range of the horizontal axis (the default range is 10), The maximum range of the vertical axis (the default range is 5); based on the key semantic features of the communication information data of IBS-D patients analyzed, the data is quantified and the key semantic meanings that the current patients want to express are analyzed by combining the clustering algorithm. Clustering through these semantics combined with the clustering algorithm can analyze the main meanings expressed by IBS-D patients, thereby reducing the error problem of semantic data caused by the emotional problems of IBS-D patients. At the same time, it can also provide a semantic data foundation for subsequent feedback, facilitating data reference for doctors during the communication process;

[0073] The medical information association module, if the first characteristic medical information exists, obtains third characteristic medical information based on the association of the first characteristic medical information with the second characteristic medical information, and stores the third characteristic medical information; specifically: in, is the jth third characteristic medical information, i∈(1,N j ), j∈(1,n); store the third feature medical information including the corresponding semantics and the calculated weights corresponding to the semantics; identify abnormal physiological data and abnormal vital signs data in the monitoring data of IBS-D patients, and further screen the key semantic features of IBS-D patients through these data. This method can further reduce the error problem of key semantic features caused by the emotional impact of IBS-D symptoms on patients, provide a further data basis for subsequent data feedback, and facilitate data reference for doctors;

[0074] If the first characteristic medical information does not exist, the second characteristic medical information is stored as the third characteristic medical information;

[0075] a medical monitoring characteristic report output module, which outputs the first characteristic medical information and the third characteristic medical information to generate an IBS-D medical monitoring characteristic report if the first characteristic medical information exists; and outputs the third characteristic medical information to generate the IBS-D medical monitoring characteristic report if the first characteristic medical information does not exist;

[0076] Furthermore, the initial report of the IBS-D medical monitoring characteristic report includes one IBS-D medical abnormality information tag and K IBS-D medical conversation information tags, where K ≥ 1; each IBS-D medical conversation information tag corresponds to k pieces of the third characteristic medical information, where k ≥ 1; and the IBS-D medical abnormality information tag corresponds to the first characteristic medical information. The medical conversation information tags in the initial report are data feedback tags of IBS-D patients, such as whether there is pain or whether the frequency of diarrhea has increased. The medical conversation information tags correspond to the semantics of the third characteristic medical information. The data feedback tags and corresponding semantics are set based on expert experience.

[0077] If the first characteristic medical information is not less than the abnormal medical information threshold, the initial report outputs the IBS-D medical abnormal information label; if the sum of the k third characteristic medical information is not less than the medical conversation information threshold, the initial report outputs the corresponding IBS-D medical conversation information label;

[0078] The abnormal medical information threshold is obtained based on the average of the first characteristic medical information of historical IBS-D patients; the medical conversation information threshold is also obtained based on the average of the third characteristic medical information of historical IBS-D patients;

[0079] generating the IBS-D medical monitoring feature report based on the output IBS-D medical abnormality information tag and the output IBS-D medical conversation information tag;

[0080] By establishing an initial report, all feedback problem data from IBS-D patients can be included to the greatest extent possible. At the same time, the feedback problem data can be screened based on the abnormal data identified in the monitoring data and the key semantic features after analysis. This method can maximize the applicability of the monitoring data and reduce the limitations of data use. At the same time, it can also match the quantified data of the key semantic features to the corresponding feature problems. The error generated by the communication information data can be reduced by converting the key semantic feature data, facilitating data feedback and providing doctors with accurate reference data.

[0081] Example 2

[0082] In order to specifically illustrate the accuracy analysis process of the abnormal irritable bowel physiological characteristics model of the irritable bowel feedback analysis model, the following content is combined for explanation:

[0083] a medical information collection module for acquiring first and second medical information of an IBS-D patient; a medical information analysis module for transmitting the first medical information to an irritable bowel feedback analysis model, determining whether first characteristic medical information of the first medical information exists, and if so, extracting the first characteristic medical information and storing the data;

[0084] Furthermore, the first characteristic medical information is obtained through the abnormal irritable bowel physiological characteristic model of the irritable bowel feedback analysis model; the abnormal irritable bowel physiological characteristic model includes a physiological characteristic data processing unit, a physiological characteristic data extraction unit, a physiological characteristic abnormality calculation unit, and a physiological abnormality characteristic output unit; by analyzing the various monitored physiological data as time series data, it is possible to analyze whether the various physiological data and vital sign data monitored by IBS-D patients have data abnormalities based on the existing data, and by identifying these abnormal data and automatically feeding back, provide doctors with more clear data references, and also provide a basis for data analysis for subsequent feedback of communication information data, thereby expanding the scope of application of IBS-D patient monitoring data;

[0085] The physiological characteristic data processing unit performs data preprocessing on the first medical information; the preprocessing includes Fourier noise reduction, time window data segmentation and Z-score normalization;

[0086] The physiological feature data extraction unit extracts IBS-D medical information features from the preprocessed first medical information; the physiological feature data extraction unit includes 5 LSTM layers, 2 spatiotemporal convolution layers, 1 attention mechanism layer and 1 feature activation layer, referring to Figure 4 As shown, the LSTM layer corresponds to Figure 4 The memory layer in the model is used, the feature activation layer adopts a fully connected layer and a Softmax classification layer, the spatiotemporal convolution layer is a 1*3 one-dimensional convolution layer, and the feature activation layer adopts a relu layer; the physiological feature data extraction unit mainly outputs abnormal labels and normal labels for physiological feature data and vital signs data. In this embodiment, the abnormal label is 1 and the non-abnormal label is 0; in order to verify the accuracy of model recognition, 5 groups of data that did not participate in training were selected in conjunction with experts to simulate the real environment. The recognition quality was judged by the accuracy and precision of the abnormal labels. The accuracy rate is the proportion of abnormal data identified in all data, and the precision rate is the proportion of abnormal data with true labels in the abnormal data identified. The recognition accuracy and precision results are shown in Table 1:

[0087] Table 1 Abnormal recognition accuracy of abnormal irritable bowel physiological feature model in 5 sets of simulated data

[0088] Simulated data set Anomaly recognition accuracy Anomaly recognition accuracy SED01 90.97％ 93.77％ SED02 91.31％ 93.62％ SED03 90.58％ 92.89％ SED04 90.69％ 93.54％ SED05 91.18％ 93.43％

[0089] From Table 1, we can see that the anomaly recognition accuracy is above 90% and the recognition precision is around 93%, indicating a high recognition accuracy for abnormal data.

[0090] The physiological characteristic abnormality calculation unit determines whether the first characteristic medical information exists based on the characteristic similarity of the IBS-D medical information characteristics;

[0091] The physiological abnormality feature output unit outputs the first feature medical information and stores the data;

[0092] transmitting the second medical information to the irritable bowel feedback analysis model, performing data comparison between the second medical information and stored historical second medical information, generating second characteristic medical information and storing the data;

[0093] a medical information association module, which, if the first characteristic medical information exists, associates the second characteristic medical information with the first characteristic medical information to obtain third characteristic medical information, and stores the third characteristic medical information;

[0094] If the first characteristic medical information does not exist, the second characteristic medical information is stored as the third characteristic medical information;

[0095] The medical monitoring characteristic report output module outputs the first characteristic medical information and the third characteristic medical information to generate an IBS-D medical monitoring characteristic report if the first characteristic medical information exists; if the first characteristic medical information does not exist, outputs the third characteristic medical information to generate the IBS-D medical monitoring characteristic report.

[0096] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A terminal for dynamic monitoring of IBS-D symptoms and interaction between doctors and patients, characterized by: include: A medical information collection module, which obtains the first medical information and the second medical information of the IBS-D patient; The first medical information includes vital sign information and physiological activity information acquired by medical equipment; the second medical information includes voice communication information and text communication information of the IBS-D patient; A medical information analysis module transmits the first medical information to an irritable bowel feedback analysis model, determines whether first characteristic medical information of the first medical information exists, and if so, extracts the first characteristic medical information and stores the data; the first characteristic medical information is obtained through an abnormal irritable bowel physiological characteristic model of the irritable bowel feedback analysis model; The abnormal irritable bowel physiological characteristic model includes a physiological characteristic data processing unit, a physiological characteristic data extraction unit, a physiological characteristic abnormality calculation unit and a physiological abnormality characteristic output unit; The physiological characteristic data processing unit performs data preprocessing on the first medical information; The physiological characteristic data extraction unit extracts IBS-D medical information features from the preprocessed first medical information; the physiological characteristic data extraction unit includes 5 LSTM layers, 2 spatiotemporal convolution layers, 1 attention mechanism layer, and 1 feature activation layer; the physiological characteristic abnormality calculation unit performs feature similarity based on the IBS-D medical information features to determine whether the first feature medical information exists; The physiological abnormality feature output unit outputs the first feature medical information and stores the data; transmitting the second medical information to the irritable bowel feedback analysis model, performing data comparison between the second medical information and stored historical second medical information, generating second characteristic medical information and storing the data; the second characteristic medical information is obtained based on the medical feedback information analysis model of the irritable bowel feedback analysis model; The medical feedback information analysis model includes a medical feedback information processing unit, a medical feedback key information extraction unit and a medical feedback weight calculation and output unit; The medical feedback information processing unit performs data feature processing on the second medical information; The medical feedback key information extraction unit compares the second medical information after data feature processing with the historical second medical information to extract key semantic features, and the medical feedback key information extraction unit adopts a Transformer model; The medical feedback weight calculation and output unit performs feedback key weight calculation based on the key semantic features to generate second feature medical information; a medical information association module, which, if the first characteristic medical information exists, associates the second characteristic medical information with the first characteristic medical information to obtain third characteristic medical information, and stores the third characteristic medical information; if the first characteristic medical information does not exist, stores the second characteristic medical information as the third characteristic medical information; The medical monitoring characteristic report output module outputs the first characteristic medical information and the third characteristic medical information to generate an IBS-D medical monitoring characteristic report if the first characteristic medical information exists; if the first characteristic medical information does not exist, outputs the third characteristic medical information to generate the IBS-D medical monitoring characteristic report.

2. The IBS-D symptom dynamic monitoring and doctor-patient interaction terminal according to claim 1, characterized in that: The physiological characteristic abnormality calculation unit determines whether the first characteristic medical information exists based on the characteristic similarity of the IBS-D medical information characteristics, including: ;in, is the first feature medical information similarity, is the IBS-D medical information feature extracted from the first medical information; Cos is the cosine similarity algorithm, For the first medical information, is the abnormal medical information of IBS-D, Dtw is the dynamic time planning algorithm, and is the feature similarity weight.

3. The IBS-D symptom dynamic monitoring and doctor-patient interaction terminal according to claim 1, characterized in that: The medical feedback weight calculation and output unit calculates the feedback key weight according to the key semantic features, including: ;in, is the key weight of the feedback of the ith , is the total number of the second medical information of the j-th cluster category, , n is the total number of cluster categories; ln is the logarithmic function with e as the base, Num is the number of key semantic features; Tran is the Transformer model, The second medical information is the lth one, , M is the total number of second medical information, Km is the semantic clustering feature label data, and is the key weight parameter for feedback.

4. The IBS-D symptom dynamic monitoring and doctor-patient interaction terminal according to claim 3, characterized in that: If the first characteristic medical information exists, obtaining the third characteristic medical information by associating the second characteristic medical information with the first characteristic medical information includes: ;in, is the jth third characteristic medical information, , ; is the i-th feedback key weight; is the first feature medical information similarity; is the total number of the second medical information of the j-th cluster category.

5. The IBS-D symptom dynamic monitoring and doctor-patient interaction terminal according to claim 1, characterized in that: If the first characteristic medical information exists, the first characteristic medical information and the third characteristic medical information are output to generate an IBS-D medical monitoring characteristic report; if the first characteristic medical information does not exist, the third characteristic medical information is output to generate the IBS-D medical monitoring characteristic report, including: The initial report of the IBS-D medical monitoring feature report includes 1 IBS-D medical abnormality information label and K IBS-D medical conversation information labels. Each of the IBS-D medical dialogue information tags corresponds to k pieces of the third characteristic medical information, ; The IBS-D medical abnormality information label corresponds to the first characteristic medical information; If the first characteristic medical information is not less than the abnormal medical information threshold, the initial report outputs the IBS-D medical abnormal information label; if the sum of the k third characteristic medical information is not less than the medical conversation information threshold, the initial report outputs the corresponding IBS-D medical conversation information label; The IBS-D medical monitoring feature report is generated according to the output IBS-D medical abnormality information tag and the output IBS-D medical conversation information tag.

Images